Neurons as Models of Thought

Artificial Neural Nets (ANNs) are great computational models, but feed-forward networks, SOMs and other ANNs are not good approximations of human neurons.

At some level ANNs may have neuro-psychological validity, but there currently is very little evidence.

Complex models of human neurons exist, and we are advancing our knowledge of human neural functioning consistently.

One of the reasons these models are so good is that we can actually look at neurons in isolation. We can physically study them.

Still Neurons are not models of human thinking.

Neurons are rather simple feature detectors.

Modelling Neurons can help us understand how they work, but these models are a long way from the way we think.